Basic doubt in analog design in cadence

[tshankar501]

When I design an op amp, I notice that if I increase my (W/L) ratio of input transistors to a large extent (kkeping L to some constant value), then vds > vdsat and vgs <vth and the result browser shows that the region of operation is 3. I suppose that region 3 indicates that this is operating in subthreshold region according to the cadence manuals....

Am I allowed to operate in this region? Mathematically, increasing (W/L) in the equation

vgs = vth +sqrt((2*Id)/(kn'*(W/L)) shows that vgs-vth cannot be negative if we neglect the negative sign in the while doing sqrt.....

 

[safwatonline]

the subthereshold region follows an exponential law not the ordinary square law.
about wether u use it or not, it depends on the modeling ,if it is good then i don't see any problem, actually low power applications usually use this region for the low current and high gm

 

[tshankar501]

oh, so even when vgs <vth, we can operate in this threshold region....then its cool but is there any problem associated with the stability of the closed loop if we operate in this region....since I always encountered stability problems if i operate in this region....my phase margin would be so low or even negative if I operate in this region....

 

[nxing]

well, the stability and the operation region are two set of problem and they are not really related each other, you still need to go through the stability analysis with this operation.

 

[safwatonline]

the only problem that i am aware of i that u cann't get good gm at high frequencies

 

[wenadinho]

transistor ft at subthreshold region is much smaller than that at saturation region.

 

[tshankar501]

I designed a cascode as well as a folded cascode amplifier and for both, I observed that when I increase the (W/L) of the input transistor to a very large extent the circuit loses stability as well as the fT of the whole circuit increases by a large amount....I dont know how did it happened.....

Added after 2 minutes:

I suppose that a RHP zero occurs because of the high cap.....but i dont have the reason....if somebody had encountered, could you please explain about this....